Perspectives on State Estimation: Spot Estimates Versus Distributions
The conventional Kalman filter gives an analytical expression for the spot estimate of the states, which is the mean of the assumed Gaussian distribution. Conventional Bayesian state estimators are developed under the assumption that the mean of the posterior of the states is the ‘best' estimat...
Saved in:
Published in | IFAC Proceedings Volumes Vol. 45; no. 15; pp. 715 - 721 |
---|---|
Main Authors | , , |
Format | Journal Article |
Language | English |
Published |
2012
|
Online Access | Get full text |
Cover
Loading…
Summary: | The conventional Kalman filter gives an analytical expression for the spot estimate of the states, which is the mean of the assumed Gaussian distribution. Conventional Bayesian state estimators are developed under the assumption that the mean of the posterior of the states is the ‘best' estimate. While this can be true in the case where the posterior can be adequately approximated as a Gaussian distribution, in general it may not hold when the distribution of the posterior is non-Gaussian. In any case, the posterior distribution, whether it is Gaussian or not, contains a lot of information that is useful. This study explores the information contained in such distributions. The need for combining Bayesian state estimation with extracting information from the distribution is demonstrated in this work. |
---|---|
ISSN: | 1474-6670 |
DOI: | 10.3182/20120710-4-SG-2026.00154 |